kaushalprasadhial / facerecog

Realtime Facial recognition system using Siamese neural network

Face Recognition System

Face Recognition system using Siamese Neural network. The model is based on the FaceNet model. OpenCV implementaion has been done for realtime face detection and recognition using the network. The model uses face encodings for identifying users.

There are two versions available:

1. Jupyter Notebook: Explains things in greater details.
   
   
2. Command line program: Python script for the same. Below are some snapshots of it.
   
   
   
   

Face recognition

There are two options:

1. Realtime facial recognition: In this the model does facial recognition in realtime using the camera feed with text overlay telling whether the user is registered with the system or not.
2. Normal facial recognition: In this the camera starts for 5-6 seconds and snaps some pictures of the user, then the picture is used for recognition.
   

Options for adding a new user:

The program uses a python dictionary for mapping for users to their corresponding face encodings.

1. Using the image path: In this the full image name is provided along with the name of the person. The image should be saved in the images folder.
2. Using the camera: The webcam starts and takes picture of the user in the frame and registers his/her.
   

Controlling the recognition accuracy:

The threshold value controls the confidence with which the face is recognized, you can control it by changing the threshold parameter value in function calls.

Here is a demo gif

NOTE: The FPS is very less in the gif that is why it seems to lag everywhere.

References:

• Code for Facenet model is based on the assignment from Convolutional Neural Networks Specialization by Deeplearning.ai on Coursera.
  https://www.coursera.org/learn/convolutional-neural-networks/home/welcome
• Florian Schroff, Dmitry Kalenichenko, James Philbin (2015). FaceNet: A Unified Embedding for Face Recognition and Clustering
• Yaniv Taigman, Ming Yang, Marc'Aurelio Ranzato, Lior Wolf (2014). DeepFace: Closing the gap to human-level performance in face verification
• The pretrained model used is inspired by Victor Sy Wang's implementation and was loaded using his code: https://github.com/iwantooxxoox/Keras-OpenFace.
• A lot of inspiration from the official FaceNet github repository: https://github.com/davidsandberg/facenet